Abstract
The Japanese scallop (Patinopecten (Mizuhopecten) yessoensis) is an important commercial species in Funka Bay, Japan, where it is farmed using the hanging culture method. Our study was based on 6 years (from 2006 to 2011) of monthly in situ observations of scallop growth at Yakumo station. To produce a basic spatial distribution dataset, we developed an interpolation solution for the shortage of Chl-a concentration data available from the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite. Additionally, we integrated four-dimensional variational (4D-VAR) assimilation water temperature data from ocean general circulation models (OGCMs), with four vertical levels (6, 10, 14, and 18 m) from the sea surface. Statistical models, including generalized additive models (GAMs) and generalized linear models, were applied to in situ observation data, satellite data, and 4D-VAR data to identify the influence of environment factors (interpolated Chl-a, temperature, and depth) on the growth of scallops, and to develop a three-dimensional growth prediction model for the Japanese scallops in Funka Bay. We considered three methods to simulate the growth process of scallops (accumulation, summation, and product), and used them to select the most suitable model. All the interpolated Chl-a concentrations and 4D-VAR temperature data were verified by shipboard data. The results revealed that GAM, using an accumulation method that was based on a combination of integrated temperature, integrated log Chl-a, depth, and number of days, was best able to predict the vertical and spatial growth of the Japanese scallop. The predictions were verified by in situ observations from different depths (R2 = 0.83–0.94). From the distribution of three-dimensional predicted scallop growth maps at each depth, it was suggested that the growth of the Japanese scallop was most favourable at 6 m and least favourable at 18 m, although variations occurred in each aquaculture region in different years. These variations were probably due to the ocean environment and climate variation.
GA-BP-based Microalgae Growth Prediction Model
